One technology that is rapidly coming to the forefront for the formation of microelectronic devices and components is electrochemical deposition, which includes both electroplating and electroless plating of metal to form microelectronic features on a microelectronic workpiece. Though electrochemical deposition has long been employed as a fundamental step in fabrication of multilevel printed circuit boards, application of electrochemical deposition to fill sub-micron interconnect features is relatively recent and poses further additional problems, including the need for more stringent control of the electrolyte bath composition.
Electrochemical deposition is a complex process involving multiple ingredients in the electrolytic bath. If the electrolytic bath is to provide high-quality deposited films (blanket or patterned) on the surface of the substrate, the concentration of several of the constituents of the bath should be maintained. As such, the ability to monitor and control the composition of the bath is one of the important factors in ensuring uniform and reproducible film properties. In semiconductor and microelectronic component applications, the electronic and morphological properties of the metal films are of principal importance in determining final device performance and reliability. The stability of later microfabrication processes in the manufacturing sequence likewise frequently depends on repeatable mechanical properties, including modulus, ductility, hardness, and surface texture of the deposited material. All of these deposit properties are controlled or strongly influenced by the composition of the electrolytic bath.
Measurement and control of proprietary organic compounds that serve to modify the deposit properties through adsorption onto and desorption from the cathode surface during, for example, electroplating, are important since they affect the diffusion rate of metal cations to nucleation and growth sites. These compounds are typically delivered as multi-component packages from chemistry vendors. One of the functions of the additive packages is to influence the throwing power of the electroplating bath: the relative insensitivity of plating rate to variations in cathodic current density across the wafer or in the vicinity of surface irregularities. The throwing power of the electrolyte has an effect on the cross-wafer uniformity of deposited film thickness and the success with which ultrafine trenches and vias (holes) are filled without included seams or voids. Organic additives have also been shown to have substantial effects on mechanical film properties. Detection and quantification of these bath constituents is complicated by the fact that they are effective at very low concentrations in the electrolyte, for example, at several ppm or less.
Bath analysis for microelectronic applications is strongly driven by the need to limit variability and maintain device yields through maintenance of optimized process parameters. One method for controlling such ingredients in an electroplating bath is to make regular additions of particular ingredients based upon empirical rules established by experience. However, depletion of particular ingredients is not always constant with time or use. Consequently, the concentration of the ingredients is not actually known and the level in the bath eventually diminishes or increases to a level where it is out of the acceptable concentration range. If the additive content concentration deviates too far from the target value, the quality of the deposit suffers and the deposit may be dull in appearance and/or brittle or powdery in structure. Other possible consequences include low throwing power and/or plating folds with bad leveling. Accordingly, further evaluation techniques may be used instead of or in an addition to the more conventional empirical techniques.
One method for evaluating the quality of an electroplating bath is disclosed in Tench U.S. Pat. No. 4,132,605 (hereafter the Tench patent). In accordance with the procedures of the Tench patent, FIG. 1 of which is illustrated here, the potential of a working electrode 10 is swept through a voltammetric cycle, including a metal plating range and a metal stripping range, for at least two baths of known plating quality and an additional bath whose quality or concentration of brightener is to be evaluated. The integrated or peak current utilized during the metal stripping range is correlated with the quality of the bath of known quality. The integrated or peak current utilized to strip the metal in the bath of unknown quality is compared to the correlation and its quality evaluated. In a preferred embodiment of said patent, the potential of an inert working electrode 10 is swept by a function generator through the voltammetric cycle. An auxiliary electrode 20 immersed in the plating bath is coupled in series with a function generator and a coulometer to measure the charge from the working electrode 10 during the stripping portion of the cycle.
An improvement to the method disclosed in the Tench patent is described by Tench and White, in the J. Electrochem. Soc., “Electrochemical Science and Technology”, April, 1985, pp. 831-834 (hereafter the Tench publication). In accordance with the Tench publication, contaminant buildup in a copper plating bath affects the copper deposition rate, and thus interferes with brightener analysis. The Tench publication teaches a technique that involves sequentially pulsing the electrode between appropriate metal plating, metal stripping, cleaning, and equilibrium potentials whereby the electrode surface is maintained in a clean and reproducible state. The method set forth in that publication is in contrast to the continuous sweep cycle utilized in the above-referenced patent. Stated otherwise, whereas the process of the Tench patent involves a continuous voltammetric sweep between about −600 mV and +1,000 mV versus a working electrode and back over a period of about 1 minute, the Tench publication pulses the potential, for example at −250 mV for 2 seconds to plate, +200 mV for a time sufficient to strip, +1,600 mV to clean for several seconds, +425 mV for 5 seconds to equilibrate, all potentials referenced to a saturated Calomel electrode. After the foregoing, the cycle is repeated until the difference between successive results are within a predetermined value range, for example, within 2% of one another.
Although the procedure of the Tench publication provides some improvement over the procedure of the Tench patent, the present inventor has recognized several areas that need improvement. For example, neither of the foregoing methods suitably address the measurement of individual constituents of electrochemical baths having two or more constituents.